Chapter 4
Operant Conditioning: Skinner's Radical Behaviorism
Chapter Outline
Modified: 2008-10-06
- I. This Chapter: Radical Behaviorism, what is it?
- Skinner's definition of psychology: a science of behavior
- But, only directly observable stimuli admitted
- But, "private" stimuli are observable by self
- Rejects mentalism
- Types of Behaviorism
- Methodological--nearly all psychologists, scientific method, behavior
- Mediational--most restrictive, only accepts operationally defined units (Hull)
- Radical--less restrictive, observable behaviors can be public or private
- Functional Analysis of Behavior
- Search for events in the environment that govern behavior: stimuli, responses
- Traditional three step causal chain: E->M->R
- Skinner suggests that mental causes are "explanatory fictions" at worst and "mental way stations" at best
- Proposes E->R instead making him an environmental determinist
- Single-N Designs
- Small-N research is conducted in many areas of psychology, especially in clinical psychology and in Skinnerian behavior analysis. The number of participants varies from 1 (the smallest N) to larger numbers of participants. Dukes (1965) notes that small–N research is most appropriate for situations involving low variability between subjects or when the opportunity to research a specific situation is limited.
- Dukes, W. F. (1965). N = 1, Psychological Bulletin, 64, 74-79.
- II. Is Skinner's Radical Behaviorism Antitheoretical?
- Skinner's theory is theoretical, it demands observable data however.
- Assumptions:
- External stimuli govern behavior
- Operant conditioning accounts for much of behavior
- Experimental (or functional) analysis of behavior
- Controlling the environment will control behavior
- Respondent conditioning
- Darwin and Skinner
- Skinner sees responses as "naturally selected" by the environment
- Skinner Box (pigeon) or Operant Chamber (rat)
- Instrumental Conditioning
- Very similar to operant
- Apparatus is different
- Mazes
- Lashley Jumping Stand (see movie)
- Key is that response can only be made on a trial-by-trial basis
- Discriminative Stimuli
- SD-stimulus signaling opportunity for reinforcement, pigeon and green light
- SD -stimulus signaling the lack of opportunity for reinforcement, frown
- Discriminative stimuli are the first step in a reinforcement situation
- Or, discriminative stimulus, followed by response, followed by reinforcement
- III. Reinforcement: "anything that follows a response that changes the probability of a response's occurring again"
- A. Positive and Negative Reinforcement
- Both RAISE operant level
- Positive reinforcement raises operant level when GIVEN after the response
- Negative reinforcement raises operant level when TAKEN AWAY after response
- B. Punishment
- LOWERS operant level
- Positive punishment lowers operant level when GIVEN after the response
- Negative punishment lowers operant level when TAKEN AWAY after the response
- C. Illustration of Reinforcement and Punishment
- Reinforcement
- Positive-response followed by food/Raise hand, get $20
- Negative-response followed by shock removal/Raise hand, shock removed/Nagging
- Punishment
- Positive-response followed by shock presentation/Raise hand, get shock
- Negative-response followed by food removal/Raise hand, lose points
- D. Primary and Secondary Reinforcers
- Primary-associated with biological drives/UCSs are identical
- food, drink, pleasure, pain, disgust, relief, fear...
- Secondary-become reinforcers after association with primary (praise, stickers)
- token economies (work for points)
- Generalized-money, prestige, power, culturally-prized variables
- IV. Reinforcement Schedules: effects from how reinforcement is delivered
- A. Continuous or Intermittent Reinforcement: continuous means EVERY response is reinforced, intermittent means SOME responses are reinforced
- Interval (time-based)
- Fixed Interval (FI)
- Variable Interval (VI)
- Ratio Schedules (response-based)
- Fixed Ratio (FR)
- Variable Ratio (VR)
- Graphic
- Superstitious Schedules
- Chance pairings between response and "reinforcement"
- Skinner paired responses with reinforcement for one trial only, that response became more common
- lucky socks, rabbit's foot and test scores
- B. Effects of Different Reinforcement Schedules: cumulative recordings
- Effects of Schedules on Acquisition and Extinction
- Continuous schedules make acquisition easier but extinguish faster
- Intermittent schedules make acquisition harder but extinguish slower
- PREE-partial reinforcement extinction effect
- Name for effect described above
- Think of breaking up and phone calls
- Spontaneous Recovery
- Occurs in operant conditioning, return to conditioning situation and conditioned response is likely to re-occur.
- Think of "sex with ex"
- Extinction and Forgetting
- Extinction and forgetting are not the same phenomenon
- Extinction is the "unconditioning" of a response
- Think of combat veteran who hears a truck backfire
- Forgetting is different, it involves the passage of time (see chapter 9)
- Effects on Rate of Responding
- VR schedules yield high rates of responding (300 pecks/second in pigeons)
- C. Schedules of Reinforcement in Everyday Life
- 1. Illustration 1: continuous gold panning extinguishing after flood
- 2. Illustration 2: intermittent gold panning slowly extinguishing
- 3. Illustration 3: child's rattle never rattles, lost, ignored
- 4. Illustration 4: child's rattle rattles, keeps it
- D. Shaping: differential reinforcement of successive approximations of final response
- Makes operant conditioning go faster
- E. Chaining: backwards chaining-linking a series of responses to a FINAL reinforcement
- Setting a table
- Start with all items but one: dish, fork, spoon, knife, napkin
- Missing is: glass
- Have LD child place the four glasses, say, "You made the table!"
- Next day leave out glass and spoon
- Next day leave out glass, spoon, and fork
- Continue until child can make table from scratch (but always the same way)
- Rodent E. Lee
- Rat would perform a series of behaviors: climbing, swimming, running a wheel, and more. All were taught by backward chaining.
- F. Chaining in Human Learning: fishing, golf, speaking
- V. Fading, Generalization, and Discrimination
- Fading
- Discrimination taught gradually
- Pigeon first taught to peck a green key
- Green is turned off briefly
- Gradually, green stays off longer and longer
- Then, dim red light replaces off (no light)
- Red stays on longer and longer
- Pigeon learns discrimination, but without errors
- Also, no peak shift
- A. Illustration 1: Teaching Pigeons to Read-Fading
- Pigeon learns to peck when it sees card with "PECK" written on it
- Pigeon does not peck when it sees card with "DON'T PECK written on it
- B. Illustration 2: Deceiving Amorous Quail-Fading
- C. Relevance to Human Learning
- Generalization-very important
- Without generalization, we would have to learn EVERY new situation
- With generalization, however, learning in one situation will transfer to similar situations
- Think of taking notes in class
- Discrimination-very important too
- Think of shoot-don't shoot situation
- VI. Applications of Operant Conditioning
- A. Instructional Applications of Positive Contingencies
- Types of Reinforcements
- consumables-food and drink
- manipulatables-inviting objects
- visual and auditory-medals, praise
- social stimuli-opportunity for social interaction (coffee break)
- tokens-symbols that can be cashed in for other reinforcements (poker chips)
- Problems with definition-is definition circular?
- Premack (biography) Principle
- Reinforcement is a response, not a stimulus
- Dynamic reinforcement hierarchy, different stimuli will be favored at different times
- Watch TV
- Eat Pizza
- Go to movies
- Play video game
- Sit on porch
- Take a walk
- Mow the lawn
- Wash the dishes
- Replace light bulb
- Pay bills
- Practice piano
- Clean garage
- McDonald's Hierarchy
- Play in the playground
- Get the toy
- Eat dessert
- Eat meal
- In either example, items lower on the hierarchy can be reinforced by the opportunity to perform items higher on the hierarchy
- Also, the hierarchy is dynamic, that means it can change
- B. Applications of Aversive Consequences
- 1. The Case Against Punishment
- Does not lead to desirable behavior
- Suppresses but does not eliminate
- Arouses negative emotions-"Honey, why'd ya leave me?"
- Does not work?
- 2. Less Objectionable Forms of Punishment
- Time out-take child out of situation
- Response-cost punishment-student agrees to loss of privileges, points, tokens after engaging in a forbidden behavior
- Reprimands ("the look")
- 3. The Case for Punishment
- Physical punishment is common
- Suppression is often the goal-don't cross the busy street, ever!
- Punishment and injurious behavior-shock children when they self-injure themselves
- Race and physical punishment-research shows interaction between race and externalizing problems.
- White parents who use physical punishment tend to have children who act out at school
- Black parents who use physical punishment tend to have children who behave at school
- Punishment may be more acceptable and accepted in Black households, while in White households, punishment is less accepted and seen as a loss of parental control
- 4. Negative Reinforcement
- Not punishment!
- Escape conditioning
- Avoidance conditioning
- Preceded shock with a signal
- Animal learns to AVOID the shock
- Animal also quits showing emotional behavior
- C. Other Applications: Behavior Management
- 1. Positive Reinforcement and Punishment
- Token economies (NYC Token)
- Earn tokens for performing desired behaviors
- Cash tokens in for reinforcers
- Contingency contracts (AKA behavioral contracts)
- Written contract (often between parent and child) that specifies target behavior to eliminate and consequences if target behavior occurs
- 2. Counterconditioning
- Systematic desensitization
- Teach relaxation
- Talk client through fearful scenario
- Stop when client shows loss of relaxation
- Resume
- 3. Extinction
- Remove reinforcer
- Remember crying child in supermarket?
- Noncontingent reinforcement-provide reinforcement that maintains problem behavior on interval schedule. The idea is to make the reinforcer less desirable
- VII. Skinner's Position: An Appraisal
- A. Contributions
- Environmental determinism
- Small but active group of adherents
- Maintain the flame lit by Watson
- B. Evaluation as a Theory
- C. Some Philosophical Objections
- Can his system be extended to cognitive concepts?
- Or, can cognitive concepts better explain behavior?
- VIII. Summary
Back to Main Learning Page
